rpc.mountd: set libtirpc nonblocking mode to avoid DOS
Subsystems (2)
A message that is not in the archive (sent off-list or not yet received): [email protected]
NeilBrown
Reviewed-byQuoted 48 lines · On 05 Nov 2014 21:24:29 +0100 [email protected] wrote:
From: Bodo Stroesser <[email protected]>
Date: Thu, 09 Oct 2014 13:06:19 +0200
Subject: [nfs-utils] [PATCH 3/3] rpc.mountd: set libtirpc nonblocking mode to avoid DOSThis patch is experimental. In works fine in that it removes the vulnerability against a DOS attack. rpc.mountd can be blocked by a bad client, that sends many RPC requests but never reads the responses. This might happen intentionally or caused by a wrong network config (MTU).
The patch switches on the nonblocking mode of libtirpc. In that mode writes can block for a max. of 2 seconds. Attackers are forced to send requests slower, as libtirpc will close a connection if it finds two requests to read at the same time.
I do not know, whether setting MAXREC could cause trouble e.g. with big replies.Signed-off-by Bodo Stroesser <[email protected]>---
support/nfs/svc_create.c +10
-0@@ -49,6 +49,8 @@#ifdef HAVE_LIBTIRPC +#include <rpc/rpc_com.h> +#define SVC_CREATE_XPRT_CACHE_SIZE (8)static SVCXPRT *svc_create_xprt_cache[SVC_CREATE_XPRT_CACHE_SIZE ] = {NULL , };@@ -401,6 +403,7 @@const struct sigaction create_sigaction = { .sa_handler =SIG_IGN , }; +int maxrec =RPC_MAXDATASIZE ;unsigned int visible, up, servport;struct netconfig *nconf;void *handlep;@@ -412,6 +415,20 @@*/ (void )sigaction (SIGPIPE , &create_sigaction,NULL ); +/* +* Setting MAXREC also enables non-blocking mode for tcp connections. +* This avoids DOS attacks by a client sending many requests but never +* reading the reply: +* - if a second request already is present for reading in the socket, +* after the first request just was read, libtirpc will break the +* connection. Thus an attacker can't simply send requests as fast as
support/nfs/svc_create.c line 434
Quoted 10 lines
support/nfs/svc_create.c +7
-0+* he can without waiting for the response. +* - if the write buffer of the socket is full, the next write() will +* fail with EAGAIN. libtirpc will retry the write in a loop for max. +* 2 seconds. If write still fails, the connection will be closed. +*/ +rpc_control (RPC_SVC_CONNMAXREC_SET , &maxrec); + handlep =setnetconfig ();if (handlep ==NULL ) { xlog(L_ERROR, "Failed to access local netconfig database: %s",
RPC_MAXDATASIZE is 9000.
This number is only relevant on the receive size.
When sending, the "sendsz" passed to svc_tli_create (which default to 64K for TCP) is used for the 'record size'.
When receiving, any 'record' in a tcp connection which is larger than 9000 bytes will be rejected.
No message to mountd or statd could ever be that large, so this number does not impose a problematic limit.
As far as I can tell, this patch is safe and is a clear improvement.
Thanks,
NeilBrown
Quoted 9 lines
-----Original Message-----
From: NeilBrown [mailto:[email protected]]
Sent: Wednesday, November 05, 2014 11:35 PM
To: Strösser, Bodo
Cc: [email protected]; [email protected]
Subject: Re: [nfs-utils] [PATCH 3/3] rpc.mountd: set libtirpc nonblocking mode to avoid DOS
...
Quoted 8 lines
RPC_MAXDATASIZEis 9000.
This number is only relevant on the receive size.
When sending, the "sendsz" passed tosvc_tli_create(which default to 64K for TCP) is used for the 'record size'.
When receiving, any 'record' in a tcp connection which is larger than 9000 bytes will be rejected.
No message to mountd or statd could ever be that large, so this number does not impose a problematic limit.
Thank you for this explanation. That's what I wanted to but didn't know.
Bodo
Quoted 7 lines
As far as I can tell, this patch is safe and is a clear improvement.
Reviewed-by NeilBrown <[email protected]>Thanks,
NeilBrown